Visual Interactive Subgroup Discovery with Numerical Properties of Interest

Subgroup discovery consists in finding subsets of individuals from a given population which have distinctive collective properties with regard to one or more properties of interest. The interest of a subgroup can be objectively assessed using appropriate statistics, but it can also be evaluated by a data analyst or domain expert. In this paper we propose an approach to subgroup discovery via distribution rules (a kind of association rules with a probability distribution on the consequent) for numerical properties of interest. The objective interest of the subgroups is measured through statistical goodness of fit tests. The subjective interest of the subgroups can be assessed by the data analyst through a visual interactive subgroup browsing procedure.